U.S. patent application number 11/894511 was filed with the patent office on 2008-03-06 for method for correcting a spectral image for optical aberrations using software.
Invention is credited to Michael P. Houlne, Mihailo V. Rebec, James E. Smous.
Application Number | 20080059100 11/894511 |
Document ID | / |
Family ID | 39047978 |
Filed Date | 2008-03-06 |
United States Patent
Application |
20080059100 |
Kind Code |
A1 |
Smous; James E. ; et
al. |
March 6, 2008 |
Method for correcting a spectral image for optical aberrations
using software
Abstract
A spectral image is corrected for optical aberrations. Tissue is
exposed to a high-intensity, narrow band of light. The narrow band
of light is scattered by at least one analyte in the tissue. Raman
signals are optically collected from the scattered light. The Raman
signals are directed to a wavelength-separating device. The Raman
signals are detected as a function of intensity and wavelength to
create the spectral image. The spectral image is corrected for
optical aberrations using a software algorithm to spatially
reassign intensity. The software may be adapted to use a reference
image to make dynamic corrections. Fluorescence signals may also be
collected.
Inventors: |
Smous; James E.; (Niles,
MI) ; Rebec; Mihailo V.; (Bristol, IN) ;
Houlne; Michael P.; (Centennial, CO) |
Correspondence
Address: |
NIXON PEABODY LLP
161 N. CLARK STREET
48TH FLOOR
CHICAGO
IL
60601
US
|
Family ID: |
39047978 |
Appl. No.: |
11/894511 |
Filed: |
August 21, 2007 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
60839166 |
Aug 22, 2006 |
|
|
|
Current U.S.
Class: |
702/104 |
Current CPC
Class: |
A61B 5/14556 20130101;
G01N 21/65 20130101; A61B 5/1455 20130101; G01N 2021/6423 20130101;
A61B 5/14532 20130101; G01N 21/6456 20130101 |
Class at
Publication: |
702/104 |
International
Class: |
G06F 19/00 20060101
G06F019/00 |
Claims
1. A method for correcting a spectral image for optical
aberrations, the method comprising the acts of: exposing tissue to
a high-intensity, narrow band of light, the narrow band of light
being scattering by at least one analyte in the tissue; optically
collecting Raman signals from the scattered light; directing the
Raman signals to a wavelength-separating device; detecting the
Raman signals as a function of intensity and wavelength to create
the spectral image; and correcting the spectral image for optical
aberrations using a software algorithm to spatially reassign
intensity.
2. The method of claim 1, wherein the narrow band of light is a
monochromatic light.
3. The method of claim 2, wherein the monochromatic light is from a
laser.
4. The method of claim 1, wherein the narrow band of light is
obtained from a polychromatic light source.
5. The method of claim 1, wherein the narrow band of light has a
wavelength of from about 300 to about 5,000 nm.
6. The method of claim 5, wherein the narrow band of light has a
wavelength of from about 800 to about 1,050 nm.
7. The method of claim 1, wherein the detecting of the Raman light
includes using a multi-pixel detector.
8. The method of claim 1, wherein the wavelength-separating device
is a defraction element.
9. The method of claim 1, wherein the wavelength-separating device
is a filter.
10. A method for determining the concentration of at least one
analyte in a fluid, the method comprising the acts of: exposing
skin tissue to a high-intensity, narrow band of light, the narrow
band of light being scattering by at least one analyte in the
tissue; optically collecting the Raman signals from the scattered
light; directing the Raman signals to a wavelength-separating
device; detecting the Raman signals as a function of intensity and
wavelength to create the spectral image; correcting the spectral
image for optical aberrations using a software algorithm to
spatially reassign intensity; and determining the concentration of
the at least one analyte using the corrected spectral image.
11. The method of claim 10, wherein the analyte is glucose.
12. The method of claim 10, wherein the narrow band of light has a
wavelength of from about 300 to about 5,000 nm.
13. The method of claim 12, wherein the narrow band of light has a
wavelength of from about 800 to about 1,050 nm.
14. The method of claim 10, wherein the wavelength-separating
device is a defraction element.
15. The method of claim 10, wherein the detecting of the Raman
signals as a function of the intensity includes vertical
binning.
16. A method for correcting a spectral image for optical
aberrations using an instrument, the method comprising the acts of:
exposing tissue to a high-intensity band of light; optically
collecting an image of the tissue; dynamically correcting the image
using software to optically correct for at least one of the
instrument and tissue; and using information from the corrected
image to perform a general diagnosis.
17. A method for correcting a spectral image for optical
aberrations using an instrument, the method comprising the acts of:
exposing tissue to a high-intensity light, the light being
scattering by at least one analyte in the tissue; optically
collecting fluorescence signals from the scattered light; directing
the fluorescence signals to a wavelength-separating device;
detecting the fluorescence signals as a function of intensity and
wavelength to create the spectral image; and correcting the
spectral image for optical aberrations using a software algorithm
to spatially reassign intensity.
18. The method of claim 17, wherein the high-intensity light has a
wavelength of from about 300 to about 5,000 nm.
19. A method for correcting a spectral image for optical
aberrations, the method comprising the acts of: optically
collecting Raman signals from the scattered light; directing the
Raman signals to a wavelength-separating device; detecting the
Raman signals as a function of intensity and wavelength to create
the spectral image; and correcting the spectral image for optical
aberrations using a software algorithm to spatially reassign
intensity.
20. A method for correcting a spectral image for optical
aberrations using an instrument, the method comprising the acts of:
optically collecting fluorescence signals from the scattered light;
directing the fluorescence signals to a wavelength-separating
device; detecting the fluorescence signals as a function of
intensity and wavelength to create the spectral image; and
correcting the spectral image for optical aberrations using a
software algorithm to spatially reassign intensity.
Description
CROSS-REFERENCE TO RELATED APPLICATION
[0001] This application claims priority to U.S. Provisional
Application Ser. No. 60/893,166 filed on Aug. 22, 2006 which is
incorporated by reference in its entirety.
FIELD OF THE INVENTION
[0002] The present invention generally relates to a method of
correcting a spectral image using software. The method may be used
to assist in determining an analyte concentration.
BACKGROUND OF THE INVENTION
[0003] The quantitative determination of analytes in body fluids is
of great importance in the diagnoses and maintenance of certain
physiological abnormalities. For example, lactate, cholesterol and
bilirubin should be monitored in certain individuals. In
particular, it is important that diabetic individuals frequently
check the glucose level in their body fluids to regulate the
glucose intake in their diets. The results of such tests can be
used to determine what, if any, insulin or other medication needs
to be administered.
[0004] In some existing techniques, a lancet may be used to draw
fluid (e.g., blood) from a user. This fluid is then used with an
instrument or meter to determine an analyte concentration. It would
be desirable to eliminate the need to use a lancet, while still
accurately determining the analyte concentration. Such applications
are referred to as non-invasive techniques.
[0005] One non-invasive technique involves using Raman signals to
determine the concentration of an analyte such as glucose. There,
however, are disadvantages in existing non-invasive methods that
use Raman or other types of signals. For example, the spectral
image of an object that is imaged using a spectrometer contains
optical aberrations. Optical aberrations such as curvature may
result in overlapping spectral bands and/or poorly resolved peaks.
This may lead to erroneous results when attempting to quantify an
analyte in a complex matrix. These optical aberrations may make it
difficult to distinguish or differentiate analytes (e.g., glucose)
from other tissue and fluid components with similar characteristic
spectra. Some existing techniques have proposed hardware solutions
to correct the optical aberrations. These hardware solutions,
however, do not have the flexibility to address selected optical
aberrations. For example, existing hardware techniques are not
adapted to address curvatures that are not fixed such as those
associated with holographic or transmission grating.
[0006] It would be desirable to have a method that has flexibility
in addressing a variety of optical aberrations in spectral
images.
SUMMARY OF THE INVENTION
[0007] According to one method, a spectral image is corrected for
optical aberrations. Skin tissue or other tissue is exposed to a
high-intensity, narrow band of light. The narrow band of light is
scattered by at least one analyte in the skin tissue or other
tissue. Raman signals are optically collected from the scattered
light. The Raman signals are directed to a wavelength-separating
device. The Raman signals are detected as a function of intensity
and wavelength to create the spectral image. The spectral image is
corrected for optical aberrations using a software algorithm to
spatially reassign intensity.
[0008] According to another method, the concentration of at least
one analyte in a fluid is determined. Skin tissue or other tissue
is exposed to a high-intensity, narrow band of light. The narrow
band of light is scattered by at least one analyte in the skin
tissue or other tissue. Raman signals are optically collected from
the scattered light. The Raman signals are directed to a
wavelength-separating device. The Raman signals are detected as a
function of intensity and wavelength to create the spectral image.
The spectral image is corrected for optical aberrations using a
software algorithm to spatially reassign intensity. The
concentration of the at least one analyte is determined using the
corrected spectral image.
[0009] According to a further method, a spectral image for optical
aberrations is corrected using an instrument. Skin tissue or other
tissue is exposed to a high-intensity band of light. An image of
the skin tissue or other tissue is optically collected. The image
is dynamically corrected using software to optically correct for at
least one of the instrument and skin tissue or other tissue.
Information from the corrected image is used to perform a general
diagnosis.
[0010] According to yet another method, a spectral image for
optical aberrations is corrected using an instrument. Skin tissue
or other tissue is exposed to a high-intensity light. The light is
scattering by at least one analyte in the skin tissue or other
tissue. Fluorescence signals are optically collected from the
scattered light. The fluorescence signals are directed to a
wavelength-separating device. The fluorescence signals are detected
as a function of intensity and wavelength to create the spectral
image. The spectral image are collected for optical aberrations
using a software algorithm to spatially reassign intensity.
[0011] In another method, a spectral image is corrected for optical
aberrations. Raman signals from the scattered light are optically
collected. The Raman signals are directed to a
wavelength-separating device. The Raman signals are detected as a
function of intensity and wavelength to create the spectral image.
The spectral image is corrected for optical aberrations using a
software algorithm to spatially reassign intensity.
[0012] In yet another method, a spectral image is corrected for
optical aberrations using an instrument. Fluorescence signals from
the scattered light are optically collected. The fluorescence
signals are directed to a wavelength-separating device. The
fluorescence signals are directed as a function of intensity and
wavelength to create the spectral image. The spectral image is
corrected for optical aberrations using a software algorithm to
spatially reassign intensity.
BRIEF DESCRIPTION OF THE DRAWINGS
[0013] The foregoing and other advantages of the invention will
become apparent upon reading the following detailed description and
upon reference to the drawings.
[0014] FIG. 1 depicts a sequence of acts in producing a corrected
spectral image according to one method.
[0015] FIG. 2a depicts a detector with an uncorrected spectral
image according to one embodiment.
[0016] FIG. 2b depicts a detector with a corrected spectral image
of FIG. 2a.
[0017] FIG. 3a is representative spectra generated from an
uncorrected spectral image of a type of aberration depicted in FIG.
2a.
[0018] FIG. 3b is the representative spectra generated from a
corrected spectral image of FIG. 3a.
[0019] FIG. 4 is a sequence of apparatus that assists in creating a
spectral image that is corrected by software for optical
aberrations according to one process.
[0020] FIG. 5 depicts a subsection used in finding the apex of the
curves at the far left and far right of an image according to one
embodiment.
[0021] FIG. 6 depicts a series of reference points above and below
the apex pixel location according to one embodiment.
[0022] FIG. 7 depicts the difference in radiuses of curvature
between the left side and right side of an image.
DETAILED DESCRIPTION OF ILLUSTRATED EMBODIMENTS
[0023] The present invention is directed to a software method for
correcting optical aberrations of a spectral image in a
wavelength-separating device. By reducing or eliminating the
optical aberrations, in one method, a quantitative analysis of an
analyte can be determined in a non-invasive manner. It is highly
desirable for the software method to preserve spectral resolution
in quantifying a particular analyte.
[0024] Analytes that may be measured include glucose, lipid
profiles (e.g., cholesterol, triglycerides, LDL and HDL),
microalbumin, fructose, lactate, bilirubin, creatinine, uric acid,
potassium, sodium, chlorine, and pH. It is contemplated that other
analyte concentrations may be determined. As used within this
application, the term "concentration" refers to an analyte
concentration, activity (e.g., enzymes and electrolytes), titers
(e.g., antibodies), or any other measure used to determine the
desired analyte.
[0025] To determine the analyte concentration in a quantitative,
non-invasive manner, the optical images in one method are summed
vertically to generate a spectrum. The ability to distinguish
analytes (e.g., glucose) from other tissue and fluid components
relies on the ability to differentiate the characteristic spectra
of the analyte of interest.
[0026] According to one method, a spectral image is corrected for
optical aberrations by exposing skin tissue or other tissue to a
narrow band of light. The narrow band of light is scattered by at
least one analyte in the skin tissue or other tissue. In this
method, the Raman light from the scattered light is optically
collected. The Raman light is directed to a wavelength-separating
device. The Raman light is detected as a function of intensity and
wavelength to create the spectral image. The spectral image is
corrected for optical aberrations using a software algorithm to
spatially reassign intensity or, in other words, to reassign
intensity values in wavelength space.
[0027] In one method, the high-intensity, narrow-band light source
may come from a variety of sources. For example, the
high-intensity, narrow-band light source may come from a
monochromatic light source that is delivered in a narrow band. One
example of a monochromatic light source is a laser-diode source. It
is contemplated that other light sources may be used such as a
light-emitting diode and incoherent lamps. The light sources may be
filtered to provide a more clearly defined (i.e., narrow) band of
light. It is also contemplated that the high-intensity, narrow-band
light may be a dye laser, gas laser, ion laser or a pumped
laser.
[0028] In one embodiment, a polychromatic light source is combined
with appropriate filtering to yield a narrow band of light. Using
such appropriate filtering, the obtained narrow band of light may
be similar to monochromatic light. It is contemplated that an
incoherent light source such as a light-emitting diode (LED) or a
light bulb may yield a narrow band of light. In Raman applications,
the LEDs or light bulb would desirably use filtering. It is
contemplated that in other non-Raman applications that use LEDs or
light bulbs may not need filtering. It is contemplated that other
monochromatic or polychromatic sources of light may be used to
obtain a narrow band of light.
[0029] The wavelength of the light source may vary but is generally
from about 300 to about 10,000 nm. The light source may be an
ultraviolet light source, a near-infrared light source, an infrared
light source, or visible light source with appropriate filtering.
The light source to be used would be a high-intensity, narrow band
of light.
[0030] The Raman spectral information in one method may be
collected in the wavelength range from about 300 nm to about 12,000
nm. However, several wavelength-dependent characteristics unique to
tissue optics and to the Raman effect can significantly impact the
ability to successfully employ the Raman technique for the
non-invasive determination of analytes in tissue. For example, at
lower wavelengths, the inherent Raman signal from analytes in
tissue is relatively strong, but tissue autofluorescence is also
relatively strong, which may overwhelm and complicate detecting the
Raman signal in the tissue. Conversely, at higher wavelengths,
tissue autofluorescence and the inherent Raman signal decrease. The
choice of the light source would be made based on a balance of the
Raman signal power and the autofluorescence interference at the
wavelengths of interest for the analyte of interest. Therefore, for
glucose analysis, it is desirable to employ a high-intensity,
narrow band light source centered at or near 830 nm and collect the
Raman spectral information in the wavelength range of from above
800 nm to about 1050 nm where the strength of the Raman signal is
optimized verses the tissue autofluorescence.
[0031] The glucose-related Raman spectral information may be
collected from Raman scattered light shifted from 100 cm.sup.-1 to
10,000 cm.sup.-1 away from the light source. More specifically, the
glucose-related Raman spectral information may be collected from
Raman scattered light shifted from 100 cm.sup.-1 to 1600 cm.sup.-1
away from the light source since the strongest glucose peaks occur
at Raman shifts of about 1340 cm.sup.-1 and about 1125 cm.sup.-1.
It is contemplated that the Raman spectral information may be
collected in different ranges, especially if the analyte
concentration to be determined is not glucose.
[0032] One specific example is an 830 nm laser-diode source. One
example of a commercially available 830 nm laser-diode source is
Invictus.TM. NIR 830 nm diode laser, which is marketed by Kaiser
Optical Systems, Inc. of Ann Arbor, Mich. Another example is a
PI-ECL-830-300 diode laser, which is marketed by Process
Instruments of Salt Lake City, Utah. In one embodiment, the laser
light is delivered to the skin tissue or other tissue in about a 1
mm beam diameter. It is contemplated that other laser-diode sources
may be employed.
[0033] The high-intensity, narrow band of light may be adjusted so
that a higher resolution Raman spectrum is generated. For example,
the high-intensity narrow band of light may be limited, resulting
in less light being exposed and a higher resolution Raman signal
being obtained. By adjusting the high-intensity, narrow band of
light, the strength of the Raman signal and the exposure may be
optimized depending on the analyte of interest.
[0034] It is contemplated that different wavelength-separating
devices may be used in the present invention. Wavelength-separating
devices that may be used in the inventive methods include
defraction elements and filters. Defraction elements generally
break up the light into individual components. Filters spatially
separate groups of wavelengths. Some filters selectively allow a
desired group of wavelengths to pass therethrough, while preventing
or inhibiting undesirable wavelengths from passing therethrough.
Other filters selectively reflect a desired group of wavelengths,
while allowing undesirable wavelengths to pass therethrough. The
resultant image from a filtering embodiment is a select group of
wavelengths.
[0035] Examples of defraction elements that may be used include,
but are not limited to, holographic gratings, diffraction gratings,
optical crystals and prisms. Holographic gratings and diffraction
grating use plane-grating techniques to form its spectral images.
Examples of filters that may be used include, but are not limited
to, acousto-optical tunable filters (AOTF) and liquid crystal
tunable filters (LCTF).
[0036] The detector may be a multi-pixel detector. Examples of
multi-pixel detectors include, but are not limited to, a
charge-coupled device (CCD), a diode array or films. It is
contemplated that other multi-pixel detectors may be used. A CCD
takes the received light and displays it as a function of intensity
and wavelength. One example of a CCD includes a pixel array of 1300
rows of pixels in the vertical direction and 1340 columns of pixels
in the horizontal direction. It is contemplated that the CCD may
have a different number of rows in the vertical direction or
columns in the horizontal direction. It is contemplated that the
detector may be a movable single-pixel detector. For example, a
single-pixel detector with a movable slit may be used. In another
embodiment, the detector may be a diode array.
[0037] The software includes an algorithm that re-maps pixel values
in an aberrant optical image to match the pixel values of a correct
image of the object. In other words, the algorithm after being
implemented properly displays the original image. The algorithm
selects how the aberrant pixels are to be re-mapped by comparing
the aberrant image to an image with a non-aberrated image. An
undesired aberration occurs when lights interacts with optical
elements (e.g., lens, gratings, filters, detector, etc.). By using
a software algorithm to correct for optical aberrations, the
inventive method improves operational flexibility. For example, if
a lens is changed in the wavelength-separating device causing a
known aberration, the software algorithm could be programmed to
correct for this new aberration. The software algorithm may also
assist in addressing small changes in optical alignment.
[0038] In one specific application, a spectral image may be formed
using a vertical straight slit image. The intensity of the spectral
image in this embodiment is typically determined by vertical
binning. For chemical identification and quantitative analysis, the
slit image is summed vertically to form the spectrum. The image of
a straight, vertical slit through a holographic wavelength
separating device, for example, results in a curved image because
light rays from different positions along the length of the
vertical slit are incident on the grating at different, oblique
angles. An uncorrected spectral image using a vertical straight
slit image includes curvature. That curvature results in a poorly
resolved spectrum with significant spectral band broadening and the
loss of resolution after binding. Thus, the optical aberration to
be corrected in this embodiment is the curvature associated with
the method of using a vertical straight slit image.
[0039] In another method, the spectral image to be corrected by
software may include the use of an AOTF. When using an AOTF, the
system can operate in a number of modalities. One modality includes
individual wavelength or narrow wavelength regions being passed
through the filter and collected by a detector (e.g., a CCD). The
aberrations inherent in the AOTF and collection optics create a
poorly resolved image, especially when added together or
integrated. The software correction process is used to reconstruct
a sharply resolved image. The software-correction process may
correct for two-dimensional aberrations as well as intensity
aberrations. The second modality is the collection of multiple
single wavelength images. Those individual wavelength images are
combined together to form a spectral image. This spectral image may
be offset by a certain degree. By using a software algorithm to
correct the spectral image, the spectral image may be realigned to
prevent or reduce blurring in the image. In such a method, the
horizontal and vertical portions of the spectral image may be
corrected. Some optical aberrations that may be addressed by the
software include positioning changes, non-coherent fibers, lens
aberrations, crystal inconsistencies, intensity aberrations, and
collection-efficacy variations.
[0040] In one method, the light (e.g., Raman signals) is collected
and detected using a spectrometer. The spectrometer is a device
that collects and separates the light. The spectrometer includes
collection optic(s), a wavelength-separating device and a detector.
The spectrometer may also include a light source. The collection
optic(s) assists in collecting and directing the light (e.g., Raman
light) through the wavelength-separating device. The
wavelength-separating device separates the light into separate
wavelength components. The detector detects the light as a function
of intensity and wavelength to create the spectral image. The
spectrometer may further include focusing optic(s) that assists in
directing and focusing the Raman light exiting from the
wavelength-separating device onto the detector.
[0041] Referring to FIG. 1, an object 10 is shown being imaged
through a wavelength-separating device 14, which is a prism in this
embodiment. The Raman light is detected as a function of intensity
and wavelength to create the spectral image by a detector 16. The
detector in this embodiment is a charge-coupled device (CCD). The
spectral image depicted in the detector 16 of FIG. 1 is a desired
corrected spectral image.
[0042] As discussed above, one uncorrected aberration of a spectral
image may be characterized by curvature of an image. The curvature
of an image occurs when the image is passed through a
wavelength-dispensing device. One example of an uncorrected image
with curvature is shown on a detector in FIG. 2a. The uncorrected
image of FIG. 2a may be from a plane-grating spectrometer that uses
generally straight vertical slits.
[0043] Specifically, FIG. 2a depicts a detector 26a with a
plurality of curved line images 28a-c of the vertical slits. FIG.
2a depicts a spectral image that has not been corrected by a
software algorithm. FIG. 2b depicts a detector 26b with a plurality
of corrected line images 30a-c that are generally straight vertical
lines. The corrected line images 30a-c of FIG. 2b were corrected
using a curvature-correction software algorithm. The horizontal
axis of the detectors 26a,b depicts the wavelength components of
the Raman light, while the vertical axis depicts the measure of the
vertical slit height. The wavelength increases going from left to
right on the horizontal axis of FIG. 2b. The intensity of the light
corresponds to the quantity of the wavelength component of the
Raman light.
[0044] In one method to correct curvature, the software algorithm
begins by using a vertical center (shown as dashed lines 32a, 32b
in respective FIGS. 2a, 2b) of the curve as a reference point. The
curvature-correction algorithm shifts intensity values in pixels
above and below the dashed line 32a shown in FIG. 2a. By shifting
the intensity values in the pixels, the shifted pixels will be
aligned with the reference pixels as shown in, for example, FIG.
2b.
[0045] FIG. 3a shows a representative spectrum of a neon light
source constructed from a curved line, while FIG. 3b shows a
spectrum of the neon light source reconstructed from the corrected
line image. The spectrums of FIGS. 3a, 3b were generated from the
respective uncorrected and corrected images by vertically summing
the pixel values. The peak shapes in the uncorrected spectrum (FIG.
3a) showed significant tailing on the right side of the image and
overlapping adjacent peaks. Thus, the spectral resolution of FIG.
3a was degraded. Such a degraded image may destroy valuable
information pertaining to chemical structure and identification.
FIG. 3b, on the other hand, showed peaks in the spectrum that were
highly resolvable and, therefore, more characteristic of neon. The
corrected spectral resolution of FIG. 3b showed less degradation,
if any, as compared to the uncorrected spectral resolution of FIG.
3a.
[0046] Referring to FIG. 4, a sequence of apparatus is shown that
assists in creating a spectral image that is corrected by software
for optical aberrations according to one process. FIG. 4 includes a
laser source 60 that generates a narrow band of light. The laser
light is transmitted through a lens 62 onto skin tissue 66. The
narrow band of light is scattered by at least one analyte in the
skin tissue. A collection mirror 70 assists in collecting the
scattered light from the skin tissue 66. It is contemplated that
other mirrors may be used in collecting the scattered light such as
a parabolic mirror. The collected light is directed to a
wavelength-separating device such as an acousto-optical tunable
filter (AOTF) 76. The collected light is directed to at least one
slit 80 via a fiber bundle 82. In one embodiment, the fiber bundle
may form the slit. The fiber bundle 82 assists in directing the
collected light into the at least one slit 80. The fiber bundle is
one example of additional detector-focusing optics that assist in
directing and focusing the light from the wavelength-separating
device onto the detector. It is contemplated that other
detector-focusing optics may be used to assist in directing and
focusing the light from the wavelength-separating device onto the
detector. The light is detected by a detector as a function of
intensity and wavelength to create the spectral image. One example
of a detector is shown in FIG. 4 as a CCD 84 with grating 88. The
spectral image is corrected for optical aberrations using a
software algorithm to spatially reassign intensity.
[0047] The optical aberrations may be corrected using different
algorithms. In one method, to correct the curvature that is seen
when using a holographic grating, a neon-calibration lamp is used
to create a "feature rich" spectrum on a CCD array. The calibration
lamp typically covers the spectral range of interest and is imaged
across the entire area of the CCD. In one example, the spectral
lines from the neon source appear as a set of curved lines imaged
on the CCD such as shown in FIG. 2a. The degree of curvature,
however, is not necessarily the same across the face of the
CCD.
[0048] The image that is formed on the CCD may be broken up into
many different spectra by reading out each row of the CCD. Each of
these spectra looks generally similar in structure, but the data
may appear to be stretched or shifted due to the curvature of the
spectral lines. To correct this curvature, original data of these
individual spectra may be shifted to eliminate the curvature in
FIG. 2a.
[0049] For example, if a 1340 column by 1300 row pixel CCD were to
be used, the image may be divided into 26 horizontal spectra. Each
of the 26 horizontal spectra is created by summing or binning 50
adjacent vertical pixels so as to create a 1340.times.26 array. The
vertical binning is done to smooth the data vertically. The number
of binned pixels may be adjusted to minimize local curvature
effects. A reference spectrum is selected that lies in the vertical
center of the array (see, e.g., dashed line 32a in FIG. 2a).
Typically, the reference spectrum is positioned on the curvature
apex. This reference spectrum is not modified, but the uncorrected
spectra that is located above and below the reference spectrum is
stretched and shifted to match the relative shape of the reference
spectrum.
[0050] To stretch the spectra, a small section on each end of the
reference spectrum (see, e.g., FIG. 5 with apex pixel locations 120
and 140) is chosen as the model spectra. The location of these
model spectra is recorded. A search routine is then called to
locate the left and right model spectra in all of the uncorrected
spectra above and below the reference spectrum as shown in FIG. 6.
The locations of the left and right model spectra is recorded. From
this data, how many pixels at each end of the uncorrected spectrum
would need to be shifted to be realigned with the reference
spectrum is determined.
[0051] In one method, to determine the amount of pixel shift at
each pixel location within one uncorrected spectrum, a linear
equation is formed from the two offset values for that spectrum,
and a pixel offset value is then calculated for all pixel locations
within that spectrum. This calculation results in 26 linear
equations, one for each row of the binned spectra. Using these
equations, a second 1340.times.26 shift map is created by
calculating the pixel shift values for each pixel in the binned
array. To map the pixel shift values of the entire 1340.times.1300
array, pixel shift values are interpolated and extrapolated
vertically from the rows of the 1340.times.26 shift map. The
resultant 1340.times.1300 array is a map that contains a horizontal
shift value for each pixel in the uncorrected spectral image. By
using this map, pixels may be relocated in subsequent spectral
images such as shown in FIG. 7.
[0052] The above described method used with respect to FIGS. 5-7
creates a linear translation map and involves image correction in
only one axis. It is contemplated that by using more model areas
within the spectra, a more exact (nonlinear) map could be created
depending on the type of aberration that is being corrected. In
addition to correcting the image in only one axis, it is
contemplated that a two dimensional correction could be performed
to correct for other optical geometric distortions, such as
spherical aberration, pincushion, barrel and coma.
[0053] Additionally, the method described above with respect to
FIGS. 5-7 involves creating a pixel translation map to characterize
a static optical system. It is contemplated, however, that if the
source of the geometric distortion is changing, a
geometric-correction algorithm may be applied dynamically to every
image that is captured. This method may be accomplished by
including a reference spectrum or geometry in each uncorrected
image that is analyzed. For example, to dynamically correct a
distorted spectral image, light from a known calibration source may
be added with each spectral image. The algorithm would then search
for and correct the image based on the location of the calibration
light's spectral features.
[0054] It is contemplated that other embodiments may be used in the
inventive methods. For example, a one-dimensional spatial image
using incoherent fiber bundles where, on one end, the fibers are
arranged as a slit. In this particular application, the
software-corrected algorithm re-maps the pixels locations in the
incoherent image of a "training object" to recreate the correct
spatial distribution.
[0055] The software-corrected algorithm is not limited to
correcting optical aberrations involving slits. For example, other
size corrections of objects may be addressed in images having
different but known, refractive indices. In addition to the
software algorithms, hardware solutions may also be used in
combination with the software-corrected algorithm to create a
corrective algorithm that would be used on subsequent collected
spectral images.
[0056] In addition to using Raman signals, other signals may be
used to correct a spectral image for optical aberrations. In
another method, a non-invasive method for determining the
concentration of an analyte uses fluorescence spectral information.
Analytes that may be measured using fluorescence spectral
information include glucose, lipid profiles (e.g., cholesterol,
triglycerides, LDL and HDL), microalbumin, hemoglobin AIc, or
bilirubin. The present invention is not limited, however, to these
specific analytes and it is contemplated that other analyte
concentrations may be determined.
[0057] In another method, a non-invasive method uses fluorescence
spectral information is used to provide a diagnosis on tissue such
as skin tissue. For example, in one method, information from the
collected fluorescence signals may be used to perform a general
diagnosis. The general diagnosis may include identifying (a) the
presence of a particular analyte; (b) a particular molecule or (c)
tissue morphology. The general diagnosis can be directed to several
beneficial applications. For example, potential cancerous skin
lesions may be identified in one application. By identifying
potential cancerous cells, the tissue removal can be minimized. In
another application, the stage of cancerous cells may be
identified. In a further application, the effectiveness of cancer
photodynamic therapy may be tracked. It is contemplated that other
diagnosis may be performed using the inventive methods.
[0058] The high-intensity light to be used with the fluorescence
spectral information may be a narrow band of light, but does not
necessarily have to be a narrow band of light. The high-intensity
light source may come from a monochromatic light source or a
polychromatic light source. It is contemplated that other light
sources may be used such as a light-emitting diode (e.g., a
fluorescence molecule LED), incoherent lamps, a dye laser, gas
laser, ion laser, a pumped laser or light bulb.
[0059] The wavelength of the light source may vary but is generally
between 300 and 10,000 nm. The fluorescence spectral information in
one method may be collected in the wavelength range from about 300
nm to about 12,000 nm. It is contemplated that the fluorescence
spectral information may be collected in different ranges depending
on the analyte concentration to be determined.
[0060] The skin tissue or other tissue is exposed to a band of
light. The tissue image may be limited to a discrete area of the
skin. In another method, the tissue image may be an entire body
scan. An image of the skin tissue or other tissue is optically
collected. Examples of imaging optics that may be used include, but
are not limited to, fiber bundles, lenses and mirrors such as
discussed above with respect to FIG. 4. The image is dynamically
corrected using software to optical correct for both the instrument
and tissue. The corrected image is then used for diagnosis. The
instrument aberrations that affect the instrument are determined.
The aberrations may be related to such items as the lens and/or
collections fibers. It is contemplated that the aberrations may be
related to other items.
[0061] It is contemplated that the method of correcting a spectral
image may be used on other items than tissue. In one method, a
spectral image is corrected for optical aberrations. Raman signals
from the scattered light are optically collected. The Raman signals
are directed to a wavelength-separating device. The Raman signals
are detected as a function of intensity and wavelength to create
the spectral image. The spectral image is corrected for optical
aberrations using a software algorithm to spatially reassign
intensity.
[0062] In another method, a spectral image is corrected for optical
aberrations using an instrument. Fluorescence signals from the
scattered light are optically collected. The fluorescence signals
are directed to a wavelength-separating device. The fluorescence
signals are directed as a function of intensity and wavelength to
create the spectral image. The spectral image is corrected for
optical aberrations using a software algorithm to spatially
reassign intensity.
Process A
[0063] A method for correcting a spectral image for optical
aberrations, the method comprising the acts of:
[0064] exposing skin tissue to a high-intensity, narrow band of
light, the narrow band of light being scattering by at least one
analyte in the skin tissue;
[0065] optically collecting Raman signals from the scattered
light;
[0066] directing the Raman signals to a wavelength-separating
device;
[0067] detecting the Raman signals as a function of intensity and
wavelength to create the spectral image; and
[0068] correcting the spectral image for optical aberrations using
a software algorithm to spatially reassign intensity.
Process B
[0069] The method of process A wherein the narrow band of light is
a monochromatic light.
Process C
[0070] The method of process B wherein the monochromatic light is
from a laser.
Process D
[0071] The method of process A wherein the narrow band of light is
obtained from a polychromatic light source.
Process E
[0072] The method of process A wherein the narrow band of light has
a wavelength of from about 300 to about 5,000 nm.
Process F
[0073] The method of process E wherein the narrow band of light has
a wavelength of from about 800 to about 1,050 nm.
Process G
[0074] The method of process A wherein the detecting of the Raman
light includes using a multi-pixel detector.
Process H
[0075] The method of process G wherein the multi-pixel detector is
a charge-coupled device (CCD).
Process I
[0076] The method of process G wherein the multi-pixel detector is
a diode array.
Process J
[0077] The method of process A wherein the detecting of the Raman
light includes using a movable single detector.
Process K
[0078] The method of process A wherein the wavelength-separating
device is a defraction element.
Process L
[0079] The method of process K wherein the defraction element is a
holographic grating, diffraction grating, optical crystal or a
prism.
Process M
[0080] The method of process A wherein the wavelength-separating
device is a filter.
Process N
[0081] The method of process M wherein the filter is an
acousto-optical tunable filter (AOTF) or a liquid crystal tunable
filter (LCTF).
Process O
[0082] The method of process A wherein the optical aberrations is a
curved line.
Process P
[0083] The method of process A wherein the image is a vertical
straight slit image.
Process Q
[0084] The method of process A wherein the detecting of the Raman
signals as a function of the intensity includes vertical
binning.
Process R
[0085] A method for determining the concentration of at least one
analyte in a fluid, the method comprising the acts of:
[0086] exposing skin tissue to a high-intensity, narrow band of
light, the narrow band of light being scattering by at least one
analyte in the skin tissue;
[0087] optically collecting the Raman signals from the scattered
light;
[0088] directing the Raman signals to a wavelength-separating
device;
[0089] detecting the Raman signals as a function of intensity and
wavelength to create the spectral image;
[0090] correcting the spectral image for optical aberrations using
a software algorithm to spatially reassign intensity; and
[0091] determining the concentration of the at least one analyte
using the corrected spectral image.
Process S
[0092] The method of process R wherein the analyte is glucose.
Process T
[0093] The method of process R wherein the narrow band of light is
a monochromatic light.
Process U
[0094] The method of process T wherein the monochromatic light is
from a laser.
Process V
[0095] The method of process R wherein the narrow band of light is
obtained from a polychromatic light source.
Process W
[0096] The method of process R wherein the narrow band of light has
a wavelength of from about 300 to about 5,000 nm.
Process X
[0097] The method of process W wherein the narrow band of light has
a wavelength of from about 800 to about 1,050 nm.
Process Y
[0098] The method of process R wherein the detecting of the Raman
light includes using a multi-pixel detector.
Process Z
[0099] The method of process Y wherein the multi-pixel detector is
a charge-coupled device (CCD).
Process AA
[0100] The method of process Y wherein the multi-pixel detector is
a diode array.
Process BB
[0101] The method of process R wherein the detecting of the Raman
light includes using a movable single detector.
Process CC
[0102] The method of process R wherein the wavelength-separating
device is a defraction element.
Process DD
[0103] The method of process CC wherein the defraction element is a
holographic grating, diffraction grating, optical crystal or a
prism.
Process EE
[0104] The method of process R wherein the wavelength-separating
device is a filter.
Process FF
[0105] The method of process EE wherein the filter is an
acousto-optical tunable filter (AOTF) or a liquid crystal tunable
filter (LCTF).
Process GG
[0106] The method of process R wherein the optical aberrations is a
curved line.
Process HH
[0107] The method of process R wherein the image is a vertical
straight slit image.
Process II
[0108] The method of process R wherein the detecting of the Raman
signals as a function of the intensity includes vertical
binning.
Process JJ
[0109] A method for correcting a spectral image for optical
aberrations using an instrument, the method comprising the acts
of:
[0110] exposing skin tissue to a high-intensity band of light;
[0111] optically collecting an image of the skin tissue;
[0112] dynamically correcting the image using software to optically
correct for at least one of the instrument and skin tissue; and
[0113] using information from the corrected image to perform a
general diagnosis.
Process KK
[0114] The method of process JJ wherein the high-intensity band of
light is a monochromatic light.
Process LL
[0115] The method of process KK wherein the monochromatic light is
from a laser.
Process MM
[0116] The method of process JJ wherein the high-intensity band of
light is obtained from a polychromatic light source.
Process NN
[0117] The method of process JJ wherein the high-intensity band of
light has a wavelength of from about 300 to about 5,000 nm.
Process OO
[0118] The method of process NN wherein the high-intensity band of
light has a wavelength of from about 800 to about 1,050 nm.
Process PP
[0119] The method of process JJ wherein optically collecting an
image includes using a multi-pixel detector.
Process QQ
[0120] The method of process PP wherein the multi-pixel detector is
a charge-coupled device (CCD).
Process RR
[0121] The method of process PP wherein the multi-pixel detector is
a diode array.
Process SS
[0122] The method of process JJ wherein the detector is a movable
single detector.
Process TT
[0123] The method of process JJ wherein optically collecting an
image includes using a defraction element as a
wavelength-separating device.
Process UU
[0124] The method of process TT wherein the defraction element is a
holographic grating, diffraction grating, optical crystal or a
prism.
Process VV
[0125] The method of process JJ wherein optically collecting an
image includes using a filter as a wavelength-separating
device.
Process WW
[0126] The method of process VV wherein the filter is an
acousto-optical tunable filter (AOTF) or a liquid crystal tunable
filter (LCTF).
Process XX
[0127] A method for correcting a spectral image for optical
aberrations using an instrument, the method comprising the acts
of:
[0128] exposing skin tissue to a high-intensity light, the light
being scattering by at least one analyte in the skin tissue;
[0129] optically collecting fluorescence signals from the scattered
light;
[0130] directing the fluorescence signals to a
wavelength-separating device;
[0131] detecting the fluorescence signals as a function of
intensity and wavelength to create the spectral image; and
[0132] correcting the spectral image for optical aberrations using
a software algorithm to spatially reassign intensity.
Process YY
[0133] The method of process XX wherein the high-intensity light is
a monochromatic light.
Process ZZ
[0134] The method of process YY wherein the monochromatic light is
from a laser.
Process AAA
[0135] The method of process XX wherein the high-intensity light is
obtained from a polychromatic light source.
Process BBB
[0136] The method of process XX wherein the high-intensity light
has a wavelength of from about 300 to about 5,000 nm.
Process CCC
[0137] The method of process XX wherein the detecting of the
fluorescence light includes using a multi-pixel detector.
Process DDD
[0138] The method of process CCC wherein the multi-pixel detector
is a charge-coupled device (CCD).
Process EEE
[0139] The method of process CCC wherein the multi-pixel detector
is a diode array.
Process FFF
[0140] The method of process XX wherein the detecting of the
fluorescence light includes using a movable single detector.
Process GGG
[0141] The method of process XX wherein the wavelength-separating
device is a defraction element.
Process HHH
[0142] The method of process GGG wherein the defraction element is
a holographic grating, diffraction grating, optical crystal or a
prism.
Process III
[0143] The method of process XX wherein the wavelength-separating
device is a filter.
Process JJJ
[0144] The method of process III wherein the filter is an
acousto-optical tunable filter (AOTF) or a liquid crystal tunable
filter (LCTF).
Process KKK
[0145] The method of process XX wherein the optical aberrations is
a curved line.
Process LLL
[0146] The method of process XX wherein the image is a vertical
straight slit image.
Process MMM
[0147] The method of process XX wherein the detecting of the Raman
signals as a function of the intensity includes vertical
binning.
Process NNN
[0148] A method for correcting a spectral image for optical
aberrations, the method comprising the acts of:
[0149] exposing tissue to a high-intensity, narrow band of light,
the narrow band of light being scattering by at least one analyte
in the tissue;
[0150] optically collecting Raman signals from the scattered
light;
[0151] directing the Raman signals to a wavelength-separating
device;
[0152] detecting the Raman signals as a function of intensity and
wavelength to create the spectral image; and
[0153] correcting the spectral image for optical aberrations using
a software algorithm to spatially reassign intensity.
Process OOO
[0154] A method for determining the concentration of at least one
analyte in a fluid, the method comprising the acts of:
[0155] exposing tissue to a high-intensity, narrow band of light,
the narrow band of light being scattering by at least one analyte
in the tissue;
[0156] optically collecting the Raman signals from the scattered
light;
[0157] directing the Raman signals to a wavelength-separating
device;
[0158] detecting the Raman signals as a function of intensity and
wavelength to create the spectral image;
[0159] correcting the spectral image for optical aberrations using
a software algorithm to spatially reassign intensity; and
[0160] determining the concentration of the at least one analyte
using the corrected spectral image.
Process PPP
[0161] A method for correcting a spectral image for optical
aberrations using an instrument, the method comprising the acts
of:
[0162] exposing tissue to a high-intensity band of light;
[0163] optically collecting an image of the tissue;
[0164] dynamically correcting the image using software to optically
correct for at least one of the instrument and the tissue; and
[0165] using information from the corrected image to perform a
general diagnosis.
Process QQQ
[0166] A method for correcting a spectral image for optical
aberrations using an instrument, the method comprising the acts
of:
[0167] exposing tissue to a high-intensity light, the light being
scattering by at least one analyte in the tissue;
[0168] optically collecting fluorescence signals from the scattered
light;
[0169] directing the fluorescence signals to a
wavelength-separating device;
[0170] detecting the fluorescence signals as a function of
intensity and wavelength to create the spectral image; and
[0171] correcting the spectral image for optical aberrations using
a software algorithm to spatially reassign intensity.
Process RRR
[0172] A method for correcting a spectral image for optical
aberrations, the method comprising the acts of:
[0173] optically collecting Raman signals from the scattered
light;
[0174] directing the Raman signals to a wavelength-separating
device;
[0175] detecting the Raman signals as a function of intensity and
wavelength to create the spectral image; and
[0176] correcting the spectral image for optical aberrations using
a software algorithm to spatially reassign intensity.
Process SSS
[0177] A method for correcting a spectral image for optical
aberrations using an instrument, the method comprising the acts
of:
[0178] optically collecting fluorescence signals from the scattered
light;
[0179] directing the fluorescence signals to a
wavelength-separating device;
[0180] detecting the fluorescence signals as a function of
intensity and wavelength to create the spectral image; and
[0181] correcting the spectral image for optical aberrations using
a software algorithm to spatially reassign intensity.
[0182] While the present invention has been described with
reference to one or more particular embodiments, those skilled in
the art will recognize that many changes may be made thereto
without departing from the spirit and scope of the present
invention. Each of these embodiments, and obvious variations
thereof, is contemplated as falling within the spirit and scope of
the invention as defined by the appended claims.
* * * * *